*** Descriptive Statistics Figures (1&2)

use "SCAD-OPs_Data"

tab centralgovactor ccode
tab regionalgovactor ccode
tab securityactor ccode
tab politicallegalactor ccode
tab politicalextralegalactor ccode
tab pressactor ccode
tab laboractor ccode
tab businessactor ccode
tab educationactor ccode
tab criminalactor ccode
tab ethnicreligiousactor ccode
tab ngoactivistactor ccode
tab foreignactor ccode
tab genericcitizenactor ccode
tab genericconceptactor ccode
tab genericlocationactor ccode
tab genericviolenceactor ccode

*** Descriptive Statistics Figures (3&4)

use "SCAD-OPs_Data"

tab centralgovtarget ccode
tab regionalgovtarget ccode
tab securitytarget ccode
tab politicallegaltarget ccode
tab politicalextralegaltarget ccode
tab presstarget ccode
tab labortarget ccode
tab businesstarget ccode
tab educationtarget ccode
tab criminaltarget ccode
tab ethnicreligioustarget ccode
tab ngoactivisttarget ccode
tab foreigntarget ccode
tab genericcitizentarget ccode
tab genericconcepttarget ccode
tab genericlocationtarget ccode
tab genericviolencetarget ccode

*** Descriptive Statistics Figure 5

use "SCAD-OPs_Data"

tab year criminalactor if ccode==70
tab year genericviolenceactor if ccode==70

*** Descriptive Statistics Figure 6

use "SCAD-OPs_Data"

tab year MBactor
tab year MBtarget

*** Descriptive Statistics Figure 7

use "SCAD-OPs_Data"

tab coherenteventactor ccode

*** Multinomial Logit

drop if etype==7
drop if etype==10
mlogit SCEType i.politicallegalactor i.politicalextralegalactor i.pressactor i.laboractor i.businessactor i.educationactor i.criminalactor i.ethnicreligiousactor i.ngoactivistactor i.foreignactor i.genericcitizenactor i.genericviolenceactor haiti mexico nigeria kenya southafrica egypt polity gdppc rgovtarget cgovtarget capitalcity duration i.year, b(0) nolog

*** Multinomial Logit Figure

use "SCAD-OPs_Data"

drop if etype==7
drop if etype==10
mlogit SCEType i.politicallegalactor i.politicalextralegalactor i.pressactor i.laboractor i.businessactor i.educationactor i.criminalactor i.ethnicreligiousactor i.ngoactivistactor i.foreignactor i.genericcitizenactor i.genericviolenceactor, base(0) nolog
margins, dydx(*) post
coefplot (, keep(*:1._predict) label(strike)) (, keep(*:2._predict) label(demonstration)) (, keep(*:3._predict) label(riot)) (, keep(*:4._predict) label(violence)), swapnames xline(0) legend(rows(1))
